遗传小波网络在齿轮箱故障诊断中的应用

针对小波神经网络常用的反向传播算法普遍存在收敛速度慢,容易陷入局部极小点,网络参数的选取只能凭实验和经验来确定等缺点,提出了一种基于遗传算法优化的小波神经网络并应用于齿轮的故障诊断。仿真结果表明,该方法充分的发挥了遗传算法的全局寻优能力,小波分析的非线性逼近能力和神经网络的自学习特性,优化了系统的收敛速度和故障诊断的精度。

Application of Genetic Wavelet Neural Network in Fault Diagnosis of Gearbox

The main disadvantage of the back propagation algorithm of wavelet neural network commonly used lies in the slow convergence speed, the optimization procedure getting easily stacked into the minimal value locally and network parameter decided by experiment and experience. This paper designs a fault diagnosis method of gearbox using GA-WNN （genetic algorithm-wavelet neural network）. Simulation result proves it has global optimization capability of GA, non-linear approximation ability of wavelet and serf-learning characteristic of neural network. This method can improve the fault diagnosis accuracy and convergence rate.